Question for Sean et al?

2022-02-10 Thread Peter Abramowitsch
Hi all I've started using the mitre ZoneAnnotator am making some optimizations and necessary changes which prevent it from leaking note text into the log file (horrors!!!) But I have a question. Turns out that back in 2012 someone left raw system.outs in the mastif-zoner code which comes to cT

Re: Question for Sean et al? [EXTERNAL]

2022-02-10 Thread Finan, Sean
Hi Peter, The fastest turnaround for you would probably be seen by modifying a local copy. According to this site the source is available: http://mastif.sourceforge.net/ Sean From: Peter Abramowitsch Sent: Thursday, February 10, 2022 2:12 PM To: dev@ct

RE: neural negation model in ctakes

2022-02-10 Thread Thomas W Loehfelm
Thanks for the BERT negation classifier Tim - I tested it on a small sample of radiology notes and didn't see a big improvement in accuracy (kudos to the original ctakes pipeline!) but your framework of a FastAPI component that can loosely interface with cTAKES was really interesting. @Peter -

Fwd: neural negation model in ctakes

2022-02-10 Thread Peter Abramowitsch
-- Forwarded message - From: Peter Abramowitsch Date: Thu, Feb 10, 2022 at 6:41 PM Subject: Re: neural negation model in ctakes To: Hi Tom My changes to negex are checked into the ctakes trunk in sub-project ctakes-ytex-uima Make sure you also pick up the updated resources: a